On 6/23/09 5:19 PM, "Tim McNamara" <address@hidden> wrote:
On Jun 23, 2009, at 12:24 PM, Carl D. Sorensen wrote:
On 6/23/09 9:16 AM, "Grammostola Rosea"
<address@hidden> wrote:
Tim McNamara wrote:
On Jun 15, 2009, at 2:00 PM, Kieren MacMillan wrote:
Wol et al:
Would it be reasonable to separate the functions of putting notes on
the staff and chord names above the staff, and let the user spell
out
the chord names separately from the notes on the staff? Doing so
might really simplify this discussion and result in better
control of
the final output.
To me (but I'm not a real experienced jazz musician or lilypond
user) I
agree with this comment.
Keep things simple!?
But this facility
a) doesn't exist in LilyPond
b) would require changes to the parser, and
c) has nobody who is willing to pursue doing it.
I think I may have written my comment poorly. What I meant was
having LilyPond *not* parse <c e g b> into a Cmaj7 chord name above
the staff at all. The parser is just going to run into trouble
trying to interpret something like <e c e ges bes d> as C9b5/E
because it can't read the intent of the user, only the notes in the
bracket about which it can only make its best guess. It would
probably come up with Em7b5sus4 or something which is not the same
thing in terms of musical intent, and musical intent is what the
musician playing the piece wants to know.
I think I understood your intent. The problem is that the *only* way we
have to input chords is in formats that enter notes (either <e c' e ges bes
d> or \chordmode {c:9.5-/e}). There is *no* facility in LilyPond for
entering chords as text.
The parsing of c:9.5-/e converts that string into a set of pitches, along
with a bass and an inversion (at least I think it does; I haven't reviewed
it carefully for a while, and when I did review it I wasn't as familiar with
LilyPond as I am now).
The project that Thomas is working on is making sure that when the output of
\chordmode{c:9.5-/e} is passed to the chordnames context, it will give bag
c9b5/E in the appropriate format.
I would recommend requiring the user to write the chord names out in
a text entry format (e.g., c1:9.5-/e or something like that) *if*
they want chord names above the staff and not parsing note entry to
get chord names (if indeed LilyPond can do this at all, I've never
looked into it). This makes the most sense to me (and I hope my
intent is clearer).
Right now, the ChordNames context works much better with chords entered in
\chordmode, because it knows the root and the inversion, rather than having
to try to guess the chord.
I suspect that there won't be a lot of effort right now trying to deal with
inversions or added basses, but that may come in the future.
In my opinion, the biggest problem we currently have is that we don't always
get good chord names out of \chordmode chords. But I think Thomas will have
that fixed shortly....